646 BIRKELAND. THE NORWEGIAN AURORA POLARIS EXPEDITION, 1902 1903. 



The tail that at the same time was observed in the evening, pointing away from the sun, is also 

 in perfect accordance with our theory and experiments. 



If, in the experiment shown in fig. 219, the cathode had been bent considerably upwards an 

 arrangement that I have carried out several times a correspondingly strong pencil of rays would have 

 passed over the terrella, but in such a manner that the nearest rays would have curved themselves 

 round it. 



This condition answers to that of the comet having passed between the earth and the sun, win n 

 the greater part of the tail will become visible on the evening side, while the retarded or deflected tail 

 on the morning side will become fainter and finally fade away, as the observations showed. It is not 

 easy to see what Mr. INNES means when he says that "the angular distance from the end of one tail to 

 the other was 240". The distance referred to is perhaps that between the extreme points of the , 

 ning and evening tails. Regarding the observations made by EGINITIS and HARTMANN, the most natural 

 explanation sfeems to be that at the time of observation there has been a narrow, fan shaped tail, with 

 off-shoots to north and south, which have pointed very nearly towards the earth, the one under KM- 

 NITIS' observation a little west of the earth, and that observed by HARTMANN having swung over until it 

 pointed a little east of the earth. This fan-shaped tail with direction towards the earth, calls to mind the 

 two in-drawn tongues of light in fig. 219, which are just off-shoots from a fan-shaped mass of light 

 such as this. A calculation of the direction of the terrestrial-magnetic lines of force, looking from the 

 earth towards the place in which the comet stood at the time of observation, gives a direction ai 

 due north and south, and thus symmetric in relation to the two crescent-shaped formations observed. 



Of some other remarkable observations of the comet's tail about the 2oth May, the folloi 

 mention may be made. 



EVERSHED, in Southern India, saw the comet in the morning sky like a huge search-light. It \v,is 

 not visible while passing across the sun's disc. 



W. VAN BEMMELEN writes from Batavia: "I saw it before dawn on the i8th and igth. The tail 

 was enormous ; it rose with a high inclination to the north from the eastern horizon, like a search-light, 

 and reached by its curvature the zenith. I began watching it at 4.30 a. m., but saw no auroral display, 

 nor could I detect anything of the comet's head passing the sun." 



There are similar accounts from Aden, St. Thomas, and Malta. 



From more northerly stations, on the other hand, there has been little to relate about the comet's 

 tail or any luminosity that might have some connection with it. The time of year, the unfavorable 

 position of the moon, and the atmospheric conditions, have contributed to this result. 



Concerning light-phenomena seen in Norway, it may be mentioned that at Fredriksstad, at 10.30 

 p. m. on the igth May, a luminous band was seen in the northern sky at a height of about 45 above 

 the horizon, extending from east to west. It was narrowest in the west, and could not be seen quite 

 down to the horizon, as the sky there was too light; but as far as could be seen, the radiant band 

 pointed straight to the sun, and extended in a slight curve right across the sky to about 50 above the 

 horizon in the north-east, where it was broader and very faint. The observer did not think that tin 

 band was an auroral band, but he was inclined to connect it with the comet's tail. 



From the telegraph-office at Tana it was reported that at 3.30 a. m. on the igth May, a light 

 seen, which resembled aurora, and could not have been a gleam of sunshine. A few strokes on tin 

 operator's alarm-bell were also noticed once or twice after the light had disappeared. 



At Tjarstad, in Sweden, similar auroral arcs were seen at the same hour. Judging from their 

 position, they were probably the same arcs (see STENQUIST'S "The Light-Phenomena, May, I9io"('), p. n)- 



(') Arkiv for Matematik, Astronomi och Fysik. Stockholm, 1912. 



